Molecular docking screens libraries of small molecules against a target protein receptor to discover new ligands capable of binding and changing the activity of the protein. The method uses several approximations, yet still has had many successes. This project aims to improve and extend the method by heavy use of changing resolution during sampling. At first, when considering a pose of the ligand and protein, a rough estimate will be used, good poses during this phase will be saved. Saved poses will be sampled at finer and finer resolutions, until the exact pose with the best score is found. Additionally, resolution of the protein structure will be used in a similar way, by comparing at first many similar protein structures until eventually a single structure capable of binding the ligand in question is found with the best score. These techniques are computational, but have a direct connection with experiment in testing the observations by many methods including inhibition assays, measuring the binding affinity through isothermal titration calorimetry, and confirmation of the protein/ligand pose through x-ray crystallography. Many retrospective tests will be conducted on these improved methods and then prospective tests on model systems and then an important GPCR (G-protein coupled receptor), the beta2-adrenergic receptor. While some structures have been found, they are only bound to a strong inverse agonist that entirely shuts off the activity. Predicting the conformations of the active form bound to an agonist, or the form bound to an antagonist remain important challenges. Additionally, predicting new ligands that bind with these various efficacies remains important for pharmaceutical and biological reasons. This system serves as an excellent test case since the protein conformations remain mostly unknown and the binding site is very narrow and almost entirely buried from bulk solvent, making accurate pose sampling very important. PUBLIC HEALTH RELEVANCE: The ultimate prospective test and focus of this proposal is predicting new ligands with new efficacies for an important receptor, the beta2-adrenergic receptor. This receptor controls, among other things, smooth muscle relaxation in lung tissue, currently agonists for this receptor are used in the treatment of asthma. It is also controls smooth muscle relaxation in the uterus, and agonists are used to delay preterm labor. By searching for new ligands with new efficacies, new treatments for these diseases may be found, perhaps with fewer side effects. Overall the whole class of GPCRs, which this receptor is a member of, will benefit from increased structure-based discovery methods and understanding, with numerous public health benefits.